I am using fastq-dump to download some files from sra. It turned that .sra files was downloaded into my home directory. So, there was a "disk quota exceeded" error, and a number of .sra.cache instead of .sra files were generated. But, the downloading is continuing. I wonder if the fastq.gz files downloaded in this case are deprecated? Thanks
The cache files are downloaded by default, see e.g. here. You can (and should IMHO) disable that, see here. In any case, using fastq-dump for download is not efficient, for a suggested workflow see my answer in this thread: Method to Check Fastq Completeness after Fastq-dump
Most data are anyway backed up at the European Nucleotide Archive, where you can download them much more efficiently and directly in fastq format. See my tutorial on that matter: Fast download of FASTQ files and metadata from the European Nucleotide Archive (ENA)
If you see .sra.cache files under SRA directory, it means your download is not complete and got interrupted for some reason.
vdb-validate command will fail for the corresponding accession.
See this post for downloading SRA datasets: https://reneshbedre.github.io/blog/fqutil.html